The Cartographer's Lost Coastline: How a 1790s Atlas Illuminates Modern Digital Preservation

On a windswept coast in western Scotland, in the closing years of the 18th century, a man named John Ainslie faced a problem familiar to any digital archivist today. He was creating what would become his definitive county atlas, a work of immense accuracy and beauty. His challenge was not just surveying the land, but deciding what was permanent enough to be etched onto a copper plate.

We think of old maps as records of what was, but they are often elegantly crafted arguments about what the mapmaker believed would endure. Ainslie’s map of the Isle of Skye, finished around 1794, is a masterpiece of this balancing act. He meticulously charted clan lands, roads, and harbors. But he also chose to depict a particular, sinuous coastline at Loch Sligachan with a confidence that modern satellite imagery reveals was entirely—and wonderfully—wrong.

This isn’t simple error. Historians of cartography suggest Ainslie likely worked from a blend of earlier, inconsistent sketches and his own observations, smoothing the contradictory data into a plausible, navigable line. The jagged, fractal truth of the shoreline was lost to the limits of his tools, the pressure of publishing deadlines, and a need for the map to be useful above all. His copper plate became the definitive source, and his invented coastline was reprinted and copied for nearly a century, overwriting reality in the public record.

Ainslie's Plate and Our Pixels

We tend to view digital preservation as a battle against oblivion—saving the file before it vanishes. Ainslie’s story forces us to consider a subtler danger: the tyranny of the definitive copy. His copper plate was our first stable, mass-producible data format. Once created, it was expensive to correct. Its output became the trusted source, proliferating its own vision of the world, errors and all, into libraries and minds.

Today, our “copper plates” are file formats, database schemas, and scanning standards. When a city government flattens a complex, relational dataset into a static PDF for “preservation,” they are creating a new Ainslie coastline—a simplified, often lossy version that will become the cited source. When a web archive captures a page but fails to capture the interactive data feed that powered its map, we are left with a beautiful but hollow shell, a coastline without the tide.

The lesson from the 1790s is not that Ainslie failed, but that preservation is never neutral. It is a series of consequential choices about what is ‘real’ enough to fix in time. His atlas, errors included, is now a priceless primary source about the 18th-century mind, its priorities, and its blind spots. Similarly, our early, clumsy saves of GeoCities pages or forgotten Flash sites are not just poor records of the past web; they are perfect records of our early understanding of what the web was.

Preserving digitally, then, requires the humility of a cartographer on a misty shore. We must document not just the data, but the process—the known bugs in the capture script, the limitations of the rendering engine, the decisions about what was too dynamic or too trivial to save. We must inscribe, alongside our pixels, a margin note: “Here there be uncertainties.” For it is in that honest margin, where the solid line gives way to the dotted one, that the true and complicated shape of our history resides.

Notes & further reading

A few pages I came back to while writing this: